778 



SCIENCE 



[N. S. Vol. XLI. No. 10(55 



The greatest advantage of the laboratory- 

 experiments is that in them the experimen- 

 tal plants may be infected at any time and 

 under any conditions. The plants may 

 be kept dry or wet and under different 

 temperatures, they may be fed in differ- 

 ent ways, and the factors of growth may 

 be influenced within wide limits. Under 

 such conditions the optimum of infection 

 may be determined for different varieties. 



The results of laboratory experiments 

 frequently differ greatly from those of 

 field experiments. For instance, in Wawe- 

 low's field experiments Triticum durum, 

 T. polonicum and T. turgidum were re- 

 sistant to Erysiphe graminis, but in his 

 greenhouse experiments they became in- 

 fected with this disease. Eeed's experi- 

 •ence in this respect was similar to that of 

 ~Wawelow. It is the opinion of the writer 

 that the host plants were strongly influ- 

 enced by circumstances, but Wawelow at- 

 tributes the different results to favorable 

 conditions in the greenhouse for the de- 

 velopment of large quantities of conidia. 



Such unbalancing of the host is not in- 

 frequent and in the natural environment 

 is due to extreme weather conditions. 

 Some species of Ribes are known to be im- 

 mune to the secidium of the pine blister 

 rust {Peridermium strdbi), but these spe- 

 cies may be infected and form secidia 

 under a bell jar. In the fleld the leaves 

 are infected, this being shown by the de- 

 velopment of slight yellow patches, but 

 the secidia never appear. The same is 

 true in the case of some varieties of wheat 

 with regard to Piiccinia, according to 

 Fraser, on account of the thickness of the 

 cuticle. This partial immunity is satisfac- 

 tory for practical purposes, and while par- 

 tially immune plants suffer in a small de- 

 gree through reduction of the assimilating 

 surface, they do not increase the danger 



of spreading the rust, as they form no new 

 sources of infection. 



Although some very profitable results 

 have been obtained, as already shown, 

 from the immunity methods discussed, the 

 problem of immiinity should be solved in 

 a different way. Immunity must not be 

 regarded as the only definite point to be 

 studied. In the case of every special dis- 

 ease efforts should be made to determine 

 the causes of resistance. That inununity 

 from different diseases is due to different 

 causes is clear and the factors which de- 

 termine this must now be sought. 



The cause of immunity of wheat and 

 barley from loose smut is among the sim- 

 plest. From the investigations of Hecke 

 and Brefeld it is known that the smut 

 spores are carried by the wind to the 

 stigma and that there they germinate and 

 find their way to the ovule through the 

 pollen tubes. As is generally known, there 

 are varieties of wheat which have closed 

 flowers, which means that fertilization 

 takes place within the glumes. In such 

 cases the smut spores can not reach the 

 stigmas at the proper time, and therefore 

 infection can not take place. In this case, 

 therefore, by investigating the question of 

 flowering the problem of resistance can be 

 solved without artifleial infection. Many 

 of the intermediate stages which exist be- 

 tween immune and susceptible races may 

 be detected by close observation. In like 

 manner several races of rye show differ- 

 ent degrees of susceptibility to ergot 

 {Claviceps purpurea), the resistance be- 

 ing least in those having a long flowering 

 period. 



The channel from the calyx to the car- 

 pels is open in many varieties of pears. 

 Such varieties are susceptible to infection 

 by Fusarium putrefaciens, as Osterwalder 

 has shown. The varieties without the open 



